Project Summary/Abstract Quantitative RT-PCR is widely used for molecular diagnostics and evaluating prognosis in cancer. While detection of biomarkers (presence vs. absent) is becoming a part of routing clinical practice, the actual quantification of tumor burden, for example, in lymph nodes or serum is still in developmental phase. In order for such actual quantification to become clinically useful, it is necessary to have in place data analysis methods that would provide high accuracy and precision of relative qRT-PCR quantification of low abundance transcripts across various equipment platforms and chemistries used for qRT-PCR assay. The overall goal of this project is to develop and validate new and universal methods of efficiency-adjusted relative qRT-PCR quantification that would be universally applicable to various qRT-PCR equipment platforms and chemistries. New quantification methods will be based on novel semiparametric models for qRT-PCR kinetic data that flexibly represent amplification history using smoothing splines and incorporate the model for dynamics of qRT-PCR efficiency through the penalty defined by suitable differential equation. The proposed studies will (1) investigate the utility of Michaelis-Menten model and its extensions for describing dynamics of qRT-PCR efficiency; (2) develop semi- parametric models for qRT-PCR kinetic data incorporating the dynamics of PCR efficiency using the profiled penalty estimation approach in functional data analysis; (3) develop new universal methods for efficiency-adjusted relative qRT-PCR quantification based on the proposed semi-parametric models; (4) compare accuracy and precision of the new and established methods using simulations and wide range of kinetic qRT-PCR data publicly available and collected during the course of two large NCI sponsored studies of GUYC2C in fresh tissue and blood from colorectal cancer patients.